PhD student Yifan Huang, Dr Iman Tavakkolnia, Prof. Majid Safari, and Prof. Harald Haas have authored a groundbreaking paper that brings us one step closer to turning the vision of advanced optical wireless 3D positioning and device orientation estimation using LiFi and AI technologies into a reality.
PhD student Yifan Huang, Dr Iman Tavakkolnia, Prof. Majid Safari, and Prof. Harald Haas have authored a groundbreaking paper that brings us one step closer to turning the vision of advanced optical wireless 3D positioning and device orientation estimation using LiFi and AI technologies into a reality. This research has far-reaching implications for everyday citizens. The technology enables reliable high-precision indoor positioning for various applications, ranging from patient monitoring in healthcare facilities to autonomous parking for self-driving vehicles.
In this study, researchers introduced a new AI model designed to improve the accuracy of positioning in Optical Wireless Communication (OWC) systems. Unlike previous models that only considered one-directional setups, this new model estimates both the 3D position and orientation of a device. By refining the model to focus specifically on position, using data like signal strength and measured orientation, they achieved an impressive accuracy with an average error of less than 2 cm.
To further enhance accuracy, they accounted for common sensor noise and used a noise reduction technique, keeping the error margin around 2 cm. The study also explored how the number of access points (APs) affects accuracy, showing that even with just 4 APs, the system could maintain an average error as low as 7.25 cm. These findings offer a promising approach for highly accurate positioning in OWC systems.
The team is now working on real-time experimental test and validation of this technology in real-world environments.
Please contact us at titan-enquiries@eng.cam.ac.uk if you are interested in learning more about this work.
See the full report
